Meissne , Leonie P.; Pe e son, Sonja; Sem au, Finn Ole
A icle — Published Ve sion
I ’s no a sp in , i ’s a ma a hon: e iewing go e nmen al
R&D suppo o en i onmen al inno a ion
Jou nal o En i onmen al Planning and Managemen
P o ided in Coope a ion wi h:
Kiel Ins i u e o he Wo ld Economy – Leibniz Cen e o Resea ch on Global Economic Challenges
Sugges ed Ci a ion: Meissne , Leonie P.; Pe e son, Sonja; Sem au, Finn Ole (2024) : I ’s no a sp in ,
i ’s a ma a hon: e iewing go e nmen al R&D suppo o en i onmen al inno a ion, Jou nal o
En i onmen al Planning and Managemen , ISSN 1360-0559, Taylo & F ancis, London, Iss. La es
A icles, pp. 1-27,
h ps://doi.o g/10.1080/09640568.2024.2359442
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/306604
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h p://c ea i ecommons.o g/licenses/by/4.0/
Jou nal o En i onmen al Planning and Managemen
ISSN: (P in ) (Online) Jou nal homepage: www. and online.com/jou nals/cjep20
I ’s no a sp in , i ’s a ma a hon: e iewing
go e nmen al R&D suppo o en i onmen al
inno a ion
Leonie P. Meissne , Sonja Pe e son & Finn Ole Sem au
To ci e his a icle: Leonie P. Meissne , Sonja Pe e son & Finn Ole Sem au (08 Jul 2024): I ’s no
a sp in , i ’s a ma a hon: e iewing go e nmen al R&D suppo o en i onmen al inno a ion,
Jou nal o En i onmen al Planning and Managemen , DOI: 10.1080/09640568.2024.2359442
To link o his a icle: h ps://doi.o g/10.1080/09640568.2024.2359442
© 2024 The Au ho (s). Published by In o ma
UK Limi ed, ading as Taylo & F ancis
G oup.
Published online: 08 Jul 2024.
Submi you a icle o his jou nal
A icle iews: 486
View ela ed a icles
View C ossma k da a
Ci ing a icles: 1 View ci ing a icles
Full Te ms & Condi ions o access and use can be ound a
h ps://www. and online.com/ac ion/jou nalIn o ma ion?jou nalCode=cjep20
REVIEW ARTICLE
I ’s no a sp in , i ’s a ma a hon: e iewing go e nmen al R&D
suppo o en i onmen al inno a ion
Leonie P. Meissne , Sonja Pe e son and Finn Ole Sem au
Depa men o Inno a ion and In e na ional Compe i ion, Kiel Ins i u e o he Wo ld Economy,
Kiel, Ge many
(Recei ed 12 Oc obe 2023; inal e sion ecei ed 8 May 2024)
In a ace agains global wa ming, he wo ld mus accele a e he de elopmen and
adop ion o en i onmen al inno a ions (EIs). In his li e a u e e iew, we explo e
he ole o go e nmen s in p omo ing EIs ac oss s ages o ma u i y and assess he
po en ial o educe emissions. Theo e ical amewo ks on ma ke impe ec ions
unde line he necessi y o go e nmen al Resea ch and De elopmen (R&D)
suppo . While emission p icing emains he mos cos -e icien clima e policy, i
ails as a s and-alone ins umen o su icien ly encou age EI. O e all, he op imal
app oach is a policy mix complemen ing emission p icing wi h go e nmen al R&D
suppo . The heo e ical inding is backed by empi ical s udies on he de elopmen
and deploymen o enewable ene gies, which also show ha in es men in R&D
can e ec i ely educe emissions. The e iew concludes by dissec ing wo pi o al
policy ini ia i es, he US In la ion Reduc ion Ac and he Eu opean G een New
Deal Indus ial Plan, e alua ing hei po en ial o e ec i ely con ibu e o
deca boniza ion.
Keywo ds: g een/eco-/en i onmen al inno a ion; R&D suppo ; clima e policy;
inno a ion policy
JEL: O32; O38; Q54; Q55; Q58
1. In oduc ion
The u gency o comba clima e change heigh ens as he deadline o ne -ze o emission
a ge s is apidly app oaching. Howe e , exis ing clean echnologies a e ine icien in
achie ing emission educ ion a ge s beyond 2030 (In e na ional Ene gy Agency 2021).
Fo u he deca boniza ion and achie ing ne -ze o emissions, i equi es en i onmen al
inno a ion (EI) o b ing imma u e clean ene gy echnologies o ma ke eadiness and
de elop a sui e o no el echnologies. Despi e he impo ance o EI, g een pa en ing
ac i i y has shown a conside able downwa d end in he pas decade such ha he cu -
en inno a ion le el is deemed o be insu icien o de i e a ne -ze o economy
(Ce an es e al. 2023; P obs e al. 2021). Bu he g een ansi ion is no only a ace
agains global clima e change bu also one o ma ke powe . The In e na ional Ene gy
Co esponding au ho . E-mail: [email p o ec ed], [email p o ec ed]
ß2024 The Au ho (s). Published by In o ma UK Limi ed, ading as Taylo & F ancis G oup.
This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License (h p://c ea i ecommons.
o g/licenses/by/4.0/), which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is
p ope ly ci ed. The e ms on which his a icle has been published allow he pos ing o he Accep ed Manusc ip in a eposi o y by
he au ho (s) o wi h hei consen .
Jou nal o En i onmen al Planning and Managemen , 2024
h ps://doi.o g/10.1080/09640568.2024.2359442
Agency (IEA) o ecas s ha he clean echnology indus y will be wo h US$650 bil-
lion annually by 2030 (In e na ional Ene gy Agency 2023). Se ing ha indus y can
con ibu e o economic p ospe i y in a ne -ze o emission u u e.
The passing o he In la ion Reduc ion Ac (IRA) in he US has ca apul ed go e n-
men al suppo o esea ch and de elopmen (R&D) o EI back on he poli ical
agenda in pu sui o hi ing he ne -ze o emissions a ge by mid-cen u y as well as
cap u ing a on - ow sea in he clean ene gy ma ke . As pa o he IRA, he US
commi s US$370 billion in ax c edi s o a clean ene gy economy o empowe
Ame ican en i onmen al inno a o s (The Whi e House 2023). While he IRA migh
ha e ueled he discussion, he US is no alone in i s e o s o deca bonize i s econ-
omy while boos ing i . Simila g een indus ial policies a e obse able in o he wo ld
egions, mos no ably Eu ope’s G een Indus ial Plan embedded in he Eu opean G een
Deal wo h a ound 600 billion e o und a jus ene gy ansi ion.
1
Despi e he eno -
mous amoun s o unding alloca ed o hese e o s, he ques ion emains as o how
e ec i e inno a ion policy is in achie ing en i onmen al objec i es.
In his li e a u e e iew, we analyze he ole and impac o go e nmen al R&D
suppo o EIs o add ess clima e change and acili a e he g een ansi ion owa ds a
ne -ze o u u e. As go e nmen al R&D suppo o EIs, we conside any o m o iscal
suppo o R&D by go e nmen s ha educes he cos o EIs such as g an s, ax c ed-
i s o subsidies o capi al cos s (Ce an es e al. 2023; Fische and Newell 2008). The
e m EI is o en used synonymously wi h g een inno a ion o eco-inno a ion, which
we de ine as an inno a ion which leads o educed en i onmen al deg ada ion h ough-
ou i s li e cycle compa ed o ele an al e na i es. Mo eo e , we conside EIs ac oss
di e en s ages o ma u i y, e.g. g een pa en s which a e g an ed o in en ions ha
a e no el o he ma ke s. he adop ion o EIs ha a e no el o he i m bu a e
al eady es ablished in a sec o (e.g. Kemp and Pea son 2007).
Building on he li e a u e e iew, we show ha go e nmen s play a signi ican ole
in p omo ing R&D o incen i ize in es men in EIs due o ma ke ailu es impeding EI
ac i i ies. These ma ke ailu es include knowledge c ea ion, educed en i onmen al
deg ada ion, ne wo k ex e nali ies, he pa h dependency o inno a ion, o incomple e
in o ma ion, o name a ew (Acemoglu e al. 2012; Aghion e al. 2016; Ce an es
e al. 2023; Ja e, Newell, and S a ins 2005; Rennings 2000). While he e a e chal-
lenges and ade-o s, go e nmen al suppo o g een R&D is he mos popula cli-
ma e policy (Dabla-No is e al. 2023; Dechezlep ^
e e e al. 2022). Howe e , as a
s and-alone policy ins umen , R&D policy is cos ine icien in educing en i onmen-
al deg ada ion (Fische and Newell 2008; Popp 2006). None heless, go e nmen al
R&D suppo complemen s ca bon p icing, educing o e all emission mi iga ion cos s
because i add esses se e al ma ke ailu es ela ed o EI c ea ion and di usion
(Fische and Newell 2008; Veugele s 2012). Acco dingly, combining ca bon p icing
and R&D subsidies in a well-balanced policy mix can e ec i ely and e icien ly accel-
e a e inno a ion and mi iga ion.
Acco dingly, heo e ical mo i a ion alida es ha g een R&D e ec i ely suppo s
he ake-o o EIs. As one o he mos comp ehensi e empi ical s udies on go e n-
men al R&D and EIs, Johns one, Ha
s
ci
c, and Popp (2010) ind ha go e nmen al,
g een R&D has signi ican ly inc eased he inno a ion ac i i y o enewable ene gy
(RE). Also, he use o a policy mix by combining R&D suppo wi h demand-pull pol-
icies is impo an o os e EI ac i i ies (Lindman and S€
ode holm 2012). Addi ionally,
go e nmen al g een R&D suppo ed he capaci y expansion o RE (Polzin e al. 2015)
2L. P. Meissne e al.
due o cos sa ings om g een R&D (Klaassen e al. 2005), and hus, leads o a
dec ease in CO
2
emissions a he coun y-le el (Pa ama i e al. 2021). Fi m-le el ana-
lysis –al hough p ima ily conside ing p i a e R&D expendi u e a he han go e n-
men al R&D –show ha g een R&D educes bo h he ene gy and ca bon in ensi y o
echnologies, leading o emission educ ions (Alam e al. 2019).
The heo e ical and empi ical necessi y o go e nmen al R&D suppo EIs is ec-
ognized by go e nmen s ac oss he globe ha ha e s a ed o in oduce ex ensi e
indus ial policy packages o add ess clima e issues h ough EI –mos no ably he US
IRA and he EU G een Deal Indus ial Plan. We a gue ha he la e can success ully
accele a e he ake-o o EI since i complemen s ca bon p icing in he EU. The
IRA’s en i onmen al success is likely cons ained o he sho e m as i is designed o
a o ma u e clean echnologies and neglec s he impo ance o imma u e EIs o he
long e m. Addi ionally, he IRA combines di e en poli ical a ge s in one policy,
blu ing he lines be ween en i onmen al and indus ial policy goals o he de imen
o he o me . By eplica ing he IRA o indus ial policy easons, he EU G een Deal
Indus ial Plan isks ine iciencies in a i s -bes policy scheme. Ul ima ely, he suppo
o EI can ad ance clean echnologies om which he global communi y may bene i .
Wi h his li e a u e e iew, we con ibu e o he EI li e a u e in mul iple ways.
Fi s , a la ge body o esea ch analyzes he ole o go e nmen al R&D in p omo ing
EIs ac oss s ages o ma u i y, bu he li e a u e s eams a e widely un ela ed despi e
including insigh s ela ing o each o he . In pa icula , he ex ensi e heo e ical and
empi ical wo ks a e seldom connec ed. Bo h li e a u e s ands highligh he ole o
go e nmen al R&D suppo in suppo ing EI, bu i s ull po en ial can only be eaped
in a policy mix ha combines bo h en i onmen al and inno a ion aspec s. Second,
mos s udies ocus on he ela ionship be ween gene al R&D suppo and inno a ion
ac i i ies. We go beyond hese s udies by elici ing he en i onmen al e ec o go e n-
men al R&D suppo . In so doing, we no only echo he impo ance o go e nmen al
R&D suppo in he con ex o EI bu also p o ide insigh s on whe he and how go -
e nmen al R&D s a egies can be pa o a ui ul en i onmen al policy. Thi d, we
ake hese insigh s o add o he deba e o he ecen and powe ul go e nmen al indus-
ial policy packages –namely he IRA and he EU G een Deal Indus ial Plan –and
discuss hei abili y o os e EI and, impo an ly, con ibu e o emission educ ions.
Finally, we highligh ques ions ha ha e emained unanswe ed by ou li e a u e e iew
and p esen p ospec s o u u e esea ch di ec ions.
The pape comp ises se e al policy implica ions. Ou main policy implica ion is
ha go e nmen al R&D suppo is a c ucial pa o he en i onmen al policy package
due o he win-ma ke ailu es be ween he en i onmen and inno a ion. Acco dingly,
he success o policies depends on he exis ence o bo h an en i onmen al and inno -
a ion policy, while a s and-alone policy is insu icien . Taking his insigh o discuss
he ecen policy ac i i ies in he US and he EU, he EU G een Deal Indus ial Plan
can be a success ul complemen o he ca bon p icing scheme, while he IRA is
unlikely o achie e emission educ ions in a cos -e icien way. In addi ion, bo h policy
packages isk a o ing ma u e EIs and neglec he impo ance o nascen EIs o each
ne -ze o a ge s.
We s uc u e he li e a u e e iew as ollows: We s a by d awing on he ex ensi e
heo e ical li e a u e on he jus i ica ions o g een go e nmen al R&D suppo o
add ess he ques ion o why such suppo is needed o s imula e EIs (Sec ion 2). We
hen assess he in e play o go e nmen al R&D suppo wi h o he policy ins umen s
Jou nal o En i onmen al Planning and Managemen 3
in ligh o he di e en jus i ica ions and ocus on he impo ance o go e nmen al
R&D in an en i onmen al policy mix (Sec ion 3). Nex , we mo e away om he ques-
ions o why go e nmen al suppo o EI makes sense and how i should be gene ally
se up and u n o i s ac ual impac s conside ing he di e en s ages o he inno a ion
p ocess om in en ion o e inno a ion h ough deploymen . In his line, we i s sum-
ma ize empi ical e idence o he suppo o go e nmen al R&D on he inno a ion and
deploymen o clean echnologies (Sec ion 4) and hen in i s abili y o os e emission
educ ions (Sec ion 5). In Sec ion 6, we apply he lessons lea ned o discuss cu en
R&D policies in he EU and he US in hei abili y o no only ac as an inno a ion/
indus y policy bu also as an en i onmen al policy. Finally, we de i e conclusions
and sugges ions o u he esea ch (Sec ion 7).
2. Jus i ica ions o g een go e nmen al R&D suppo
In his sec ion, we add ess he ques ion o why go e nmen al R&D suppo o EI is
needed and explo e he nume ous jus i ica ions ound in he li e a u e. O en, hey a e
ela ed o ma ke impe ec ions o e en ma ke ailu es. To p o ide a s uc u ed o e -
iew, we g oup hem in o i e ca ego ies, e en hough we acknowledge ha hese
some imes o e lap. In doing so, we s ongly build on he ounda ions laid by Ja e,
Newell, and S a ins (2005) and amend hei discussion on ex e nali ies, e sions o
dynamic inc easing e u ns and unce ain ies by o he aspec s including pa h dependen-
cies –as highligh ed by he seminal con ibu ions o Acemoglu e al. (2012) and
Aghion e al. (2016)–and indus ial policy a ge s –as highligh ed in Rod ik (2014).
2.1. Ex e nali ies: knowledge c ea ion and educed en i onmen al deg ada ion
Knowledge is non- i al and o en non-excludable. Due o hese public good cha ac e -
is ics, inno a i e i ms canno ully in e nalize he gains o inno a ions (G ossman
and Helpman 1991). The di usion o knowledge o o he ma ke pa icipa ions can
be signi ican . Fo US i ms, Mye s and Lanahan (2022) ind ha e e y go e nmen -
suppo ed g an esul ed in spillo e e ec s leading o h ee mo e pa en s by o he s.
Al hough he s udy is no limi ed o g een pa en s, Rod ik (2014) a gues ha he no -
el y o EIs, he highly expe imen al na u e, and he isks o pionee ing en ep eneu s
a e cha ac e is ics making EIs p one o he ma ke ailu e o non-in e nalized know-
ledge spillo e s. In addi ion, EIs a e cha ac e ized by educed en i onmen al deg ad-
a ion, e.g. deca boniza ion. In combina ion, he posi i e knowledge c ea ion and
educed en i onmen al deg ada ion lead o an unde in es men o i ms in EIs, which
is known as he double ex e nali y p oblem (Ja e, Newell, and S a ins 2005; Popp
2006; Rennings 2000). EI policy is hus pa o an op imal se o public policies o
incen i ize g een in es men . Such an EI policy is ele an o he in en ion and di u-
sion phase o EIs. A lack o go e nmen al policies esul s in less in es men , as i
would be socially desi able (Ja e, Newell, and S a ins 2005). Howe e , inding an
op imal R&D subsidy o in e nalize ex e nali ies is no an easy ask. The main eason
is he in e empo al dimension o EIs induced by dynamic inc eased e u ns (Lancke
and Quaas 2019).
4L. P. Meissne e al.
2.2. Dynamic inc easing e u ns: lea ning by using, lea ning by doing and ne wo k
ex e nali ies
Adop ion ex e nali ies desc ibe ha he cos s o using a pa icula echnology depend
on he numbe o use s who ha e al eady adop ed i and he p oduc ion o he good
i sel . Such dynamic inc easing e u ns can be gene a ed by lea ning by using, lea ning
by doing and ne wo k ex e nali ies (Ja e, Newell, and S a ins 2005).
Lea ning by using e e s o he lea ning p ocess ha occu s when o he s obse e
he applica ion o a new echnology. Consequen ly, he adop e o EI c ea es a posi i e
ex e nali y by gene a ing in o ma ion abou he exis ence, cha ac e is ics, and success
o he new echnology (Ja e, Newell, and S a ins 2005). Lea ning by doing sheds
ligh on he supply-side o he EI adop ion. Wi h p oduc ion expe ience, cos s end o
all signi ican ly (Ja e, Newell, and S a ins 2005). Goulde and Ma hai (2000) dis in-
guish be ween R&D-based and lea ning by doing-based knowledge c ea ion. R&D-
based knowledge c ea ion lowe s he ma ginal cos s o aba emen in he u u e bu
inc eases he cos s o aba emen oday ela i e o he u u e. Simila ly, lea ning by
doing-based knowledge c ea ion a ec s ma ginal cos s in he p esen and u u e. In
addi ion, aba emen oday lowe s he cos s o aba emen in he u u e. The lea ning
a e desc ibes he educ ion o cos o each doubling o cumula i e p oduc ion o cap-
aci y and de ines he so-called lea ning cu e ha links, e.g. he cumula i e p oduc ion
o a echnology o i s cos s. P oduc i i y gains ob ained by lea ning by doing a e also
a jus i ica ion o go e nmen s o o e -p opo ionally subsidize cos lie echnologies
because hese cos s migh all conside ably in he long e m (Lancke and Quaas
2019). Al eady in 2006, Neme (2006) emphasizes he his o ically unique speed o
echnology de elopmen obse able in sola pho o ol aics. Howe e , he inds ha
lea ning by doing is only weakly d i ing he decline in p oduc ion cos s. Ins ead, a
b oade se o in luences, such as echnical ba ie s, indus y s uc u e and cha ac e is-
ics o demand a e ele an d i e s explaining he decline. Howe e , Lindman and
S€
ode holm (2012) emphasize he geog aphical domain o lea ning and s ess ha s ud-
ies allowing o he p esence o global lea ning ind highe lea ning a es. By sys em-
a ically analyzing he li e a u e on he lea ning a es epo ed o 11 powe -gene a ing
echnologies, Rubin e al. (2015) show ha lea ning can be powe ul. Fo ins ance, o
onsho e wind, hey ind an a e age lea ning a e o 12% and o sola pho o ol aic
ene gy sys ems a lea ning a e o e en 23%.
Finally, ne wo k ex e nali ies desc ibe ha he alue o echnologies inc eases wi h
he numbe o use s (Be nd , Pindyck, and Azoulay 2003; Ja e, Newell, and S a ins
2003). This dynamic can lead o a dominance o echnology, e en hough close subs i-
u es a e a ailable (Aghion e al. 2019; Be nd , Pindyck, and Azoulay 2003). Such a
lock-in –also known as pa h dependency –is obse able o some ossil echnologies
and is a majo cons ain hampe ing he ma ke -based ake-o o EIs (Ce an es e al.
2023). The dynamic na u e and pa h dependency o EIs make i mo e challenging o
design op imal R&D suppo ha emains op imal in he long e m (Acemoglu e al.
2016; Lancke and Quaas 2019). We discuss he pa h dependency o inno a ion in
mo e de ail nex .
2.3. Pa h dependency o inno a ion
Se e al s udies show ha pa h dependencies do exis o bo h clean and di y p oduc-
ion. Aghion e al. (2016) empi ically e eal o he au omo i e sec o ha egions and
Jou nal o En i onmen al Planning and Managemen 5
i ms wi h a specializa ion in di y pa en ing, show lowe ac i i ies in g een pa en ing
in he u u e. They also ind a pa h dependency in clean echnologies: i ms’his o y in
g een pa en ing de e mines he likelihood o u u e g een pa en ing.
Acemoglu e al. (2012) emphasize ha a oiding a echnological lock-in in di y
p oduc ion calls o go e nmen al ac ion. Go e nmen s ha e a c ucial ole in p e en -
ing he economy om heading owa ds an en i onmen al disas e due o he pa h
dependency o di y echnologies. Wi hou in e en ion, inno a ion and p oduc ion
would be di ec ed o di y sec o s because hese sec o s ha e a compa able ad an age
agains clean echnologies. Fi s , a ma ke size e ec di ec s inno a ion owa ds he
sec o s wi h la ge inpu ma ke s, e.g. in he ma ke o es ablished echnologies.
Fu he mo e, scien is s build on he exis ing s ock o knowledge and di ec hei
esea ch o a eas ha a e well unded and whe e o he expe ienced scien is s a e wo k-
ing. These scien is s can build hei esea ch on he ideas and knowledge o ‘s and on
he shoulde s o gian s’. Second, a p ice e ec di ec s inno a ion owa ds sec o s wi h
highe p ices, which is na u ally he ela i ely pollu ing sec o .
Aghion e al. (2019) discuss u he sou ces o a pa h dependency di ec ed
owa ds di y echnologies. Fi s , he e is a ne wo k e ec because o incen i es o
deploy inno a ions ha use exis ing in as uc u e, e.g. cha ging s a ions o elec ic
ehicles s. pe ol s a ions o ca s wi h a combus ion engine o sma g ids a e he
ounda ion o sma me e s. B eaking pa h dependencies equi es swi ching cos s,
which p i a e ac o s migh no be willing o pay. Second, especially in he ini ial
phase, shi ing o a g een economy ies up p oduc ion ac o s, which po en ially
es ic s d i e s o long- e m economic g ow h. Thi d, di e en echnologies un old a
highe payo as complemen s, e.g. enewable ene gies show a highe payo comple-
men ed by s o age capaci ies.
No ably, Lancke and Quaas (2019) model he op imal subsidy o in e nalize ex e -
nali ies, while conside ing he in e empo al dimensions o EIs induced by pa h
dependency and lea ning by doing. They ind ha he op imal subsidy should conside
he ini ial p oduc i i y o a echnology. As a esul , subsidies should be highe o less
ad anced echnologies, p o iding incen i es o echnology di e si ica ion. The
app oach is deemed op imal when p oduc i e si es a e sca ce, limi ing u u e know-
ledge u iliza ion, and when echnologies ma u e apidly wi h limi ed po en ial o u -
he lea ning.
2.4. Incomple e in o ma ion and inancial cons ains
Unce ain ies ega ding in es men cos s and he e u ns on inno a ion a e an add-
i ional domain o ma ke impe ec ions (Ja e, Newell, and S a ins 2005). While inno-
a o s ha e a mo e comp ehensi e unde s anding o he isks and oppo uni ies o new
g een echnologies, in es o s ace incomple e in o ma ion, leading hem o demand a
isk p emium o compensa e o such unce ain y. As a esul , he e is less p i a e
R&D ac i i y han socially desi able. Such a pa e n, o ins ance, pa ly explains
unde in es men in ene gy-sa ing echnologies, such as hose ela ed o housing.
House owne s may be hesi an o in es in ene gy-sa ing echnologies i hey ha e
unce ain ies abou he magni ude o sa ings in hei ene gy bill, which can ul ima ely
esul in educed, o e en non-exis en , in es men s in ene gy sa ings (Ja e, Newell,
and S a ins 2005).
6L. P. Meissne e al.
In gene al, Bond, Ha ho , and Van Reenen (2005) empi ically show ha inancial
cons ain s signi ican ly discou age in es men s in R&D. A lack o access o ex e nal
unds hinde s especially young and small companies o inno a e. This can educe he
pace o g een ans o ma ion because new companies a e ypically he companies ha
inno a e adically, while olde companies ocus on inc emen al changes. Ven u e cap-
i al is a ehicle o enable g ea e isk- aking and o un old he inno a i e capaci y o
hese companies (Ce an es e al. 2023). Go e nmen s can suppo en u e capi al by
di e en means such as ax b eaks o bene icial egula ions o unds o in es in
espec i e s a ups and small companies showing high g ow h a es. Financial con-
s ain s migh also be an issue o p i a e households.
2.5. Accep abili y and g een indus ial policy
As ou lined so a , wi hou go e nmen al in e en ion he e is an unde in es men in
g een R&D in he p i a e sec o . Rod ik (2014) discusses why many economis s a e
adi ionally eluc an o a o g een R&D policy. Fi s , he capabili y o policymake s
in achie ing well- a ge ed and e ec i e in e en ions is ques ioned. Second, he jus i i-
ca ions o go e nmen al ac ion discussed so a a e alid om he pe spec i e o a
decision make aiming o imp o e global wel a e. Howe e , o en i onmen al deg ad-
a ion, whe e he damage is global and no locally es ic ed, such jus i ica ions a e no
necessa ily binding o na ional go e nmen s a ge ing domes ic wel a e. Thi d, know-
ledge ex e nali ies o R&D a e equen ly global a he han na ional. In an in e con-
nec ed wo ld, knowledge and lea ning apidly spill o e bo de s, e.g. along global
alue chains (e.g. De Loecke 2007; Hanley and Sem au 2022; Sem au 2023)o
be ween di e en a ilia ions o mul ina ional companies (e.g. B ucal, Ja o cik, and
Lo e 2019; Kannen, Sem au, and S eglich 2021). The in e na ional di usion o EIs
opens oppo uni ies o global clima e ac ion because mos g een R&D ac i i ies ake
place in indus ialized coun ies while he bulk o emission g ow h is happening in
eme ging coun ies (Ce an es e al. 2023; Copeland, Shapi o, and Sco Taylo 2022).
Howe e , go e nmen s an icipa ing such spillo e s migh be eluc an o inancially
suppo g een R&D (Rod ik 2014).
Al hough he bene i s o g een R&D policy a e o en a he global le el a he han
he na ional le el, g een R&D suppo is a popula policy ool a ound he wo ld. The
popula i y can be explained by he ac ha o he clima e policies, such as emission
p icing, ha e dis ibu ional consequences o business models in di y echnologies and
al e na i es ha e li le poli ical appeal once hey isk educing economic ac i i ies
(Fische and Newell 2008). In line wi h his, Dechezlep ^
e e e al. (2022) ind, in a
c oss-coun y su ey, ha g een R&D suppo schemes a e mo e popula among o e s
and ci izens han al e na i es, such as ca bon p icing, bans o egula ion. Simila ly,
Dabla-No is e al. (2023) show ha ca bon p icing is a ela i ely unpopula policy
ins umen . Howe e , hey s ess ha using e enues o suppo g een in as uc u e
and low-ca bon echnologies can inc ease public accep abili y. Fu he mo e, Rod ik
(2014) s a es suppo o he domes ic indus y in global compe i ion is he main eason
o go e nmen s o subsidize g een R&D. G een indus ial policy can po en ially c ea e
a i s -mo e ad an age by edi ec ing economic ac i i ies owa ds clean echnologies
and enabling long- e m compa a i e ad an ages. Howe e , he in en ion o shi en s
om o eign p oduce s o domes ic p oduce s a ge ed o c ea e na ional
Jou nal o En i onmen al Planning and Managemen 7
emissions in China, inding ha an addi ional pa en applica ion educes a i m’sSO
2
emissions by 2.7%.
While he s udies p ima ily conside ed p i a e R&D expendi u e a he han go -
e nmen al R&D suppo , hese s udies indica e ha R&D in es men s lead o emissions
educ ions –e en when subjec ed o an emissions cap. This holds in a ious coun ies
and is mos p onounced in hea y emi ing indus ies. Ne e heless, whe he such en i -
onmen al imp o emen s can hold a he mac o-le el will be explo ed subsequen ly.
5.2. Coun y-le el e ec s o R&D suppo on CO
2
emissions
A second se o empi ical li e a u e analyses he impac o g een R&D on o e all
emissions and hus i s e ec i eness as a clima e policy a he han an inno a ion pol-
icy. Wang e al. (2012) explo e he nexus be ween ene gy echnology pa en s and CO
2
emissions in China. Examining ca bon- ee and ossil ene gy echnology pa en s, hey
show ha only an inc ease in ca bon- ee ene gy echnology pa en s educes CO
2
emissions (ac oss all Chinese egions and a he na ional le el) bu no an inc ease in
ossil- uel ene gy echnology pa en s. Thus, i is no he ene gy e iciency pa h ha
leads o emission educ ions a he mac o-le el bu a ge ed R&D o ca bon- ee
ene gy echnologies. Ne e heless, Pa ama i e al. (2021) show, in hei panel es ima-
ion o EU coun ies, ha educ ions in CO
2
emissions only depend o a small ex en
on inc easing RE consump ion. They ind ha a 1% inc ease in R&D expendi u e
(bo h public and p i a e) leads o a 0.41% inc ease in RE consump ion –which is in
line wi h he esea ch by Polzin e al. (2015)–and a dec ease o 0.14% in CO
2
emis-
sions. Meanwhile, a 1% inc ease in RE consump ion educes CO
2
emissions by 0.2%
and hus, explains a ound 0.11% o he educ ion in CO
2
emissions om g een R&D.
The e o e, he expansion o RE canno be he sole con ibu o o emission educ ions
and o he ac o s, such as imp o emen s in e iciency o inc emen al inno a ions, con-
ibu e o he educ ion in CO
2
emissions. Replica ing he s udy o OECD coun ies,
Alam e al. (2021) unde line he signi ican nega i e e ec o R&D expendi u e on
CO
2
emissions, wi h a 1% inc ease in R&D expendi u e leading o a 0.25% educ ion
in CO
2
emissions. Thus, while a educ ion in CO
2
emissions om (g een) R&D can
be ound, i is ha d o disce n i s d i e s.
Ano he ele an s and o li e a u e ela ed o he nexus be ween inno a ion and
CO
2
emissions is conce ned wi h he En i onmen al Kuzne s Cu e (EKC) (Mensah
e al. 2018). This builds on Kuzne s (1955) who hypo hesized ha he ela ionship
be ween income inequali y and economic g ow h is cha ac e ized by an in e ed U-
shape: A he beginning, emissions inc ease wi h g ow h bu wi h economic g ow h
comes he oppo uni y o inno a e and decouple emissions om economic g ow h and
dec ease. The e o e, R&D in es men s play an essen ial ole in deca bonizing he
economy.
S udying pa en da a in OECD coun ies, Mensah e al. (2018) ind a signi ican ,
nega i e ela ionship be ween CO
2
emissions and inno a ion h ough pa en s in a ew
OECD coun ies and, ul ima ely, could only pa ially p o e he alidi y o he EKC.
Fo Wes Asian and Middle Eas coun ies, Kihombo e al. (2021) show ha , while
inancial de elopmen con ibu es o en i onmen al deg ada ion, R&D mi iga es emis-
sions and hus a oids en i onmen al deg ada ion. In he mos ex ensi e esea ch
endea o , Shahbaz e al. (2020) use his o ical da a om 1870 o 2017 o s udy he
impac o economic g ow h and R&D expendi u e on UK emissions in he sho and
14 L. P. Meissne e al.
e y long e m. They ind ha he ela ionship be ween R&D expendi u e and emis-
sions can indeed be ep esen ed by an in e ed U-shape, as hypo hesized by he EKC.
Ul ima ely, bo h i m and coun y-le el analyses highligh ha (g een) R&D
educes CO
2
emissions, showing ha R&D expendi u e –whe he p i a e o go e n-
men al –is an e ec i e ool o encou age deca boniza ion. While a he i m-le el
R&D expendi u e is shown o be e ec i e in he sho e m, he educ ion in na ional
CO
2
emissions h ough R&D is mos likely a long- e m p ocess. Since he in es men s
in g een R&D and he expansion o RE lead o ac ual emissions educ ions, he
ebound e ec does no coun e ac he ull e ec o go e nmen al R&D suppo on
CO
2
emissions. Ne e heless, hese s udies examine R&D in gene al and do no di e -
en ia e public and go e nmen al R&D. Acco dingly, i canno be disce ned o wha
ex en go e nmen al g een R&D plays a ole in emission educ ion bu om he sign
o he e ec s, he s udies imply ha inc eased go e nmen al suppo o g een R&D
likely con ibu es o emission educ ions.
6. Compa ing he US Indus ial In la ion Reduc ion Ac and he EU G een Deal
Combining bo h he heo e ical and empi ical lessons on he ole o (g een) R&D sup-
po , we analyze cu en g een R&D schemes implemen ed o boos en i onmen al
inno a ion. In many coun ies, go e nmen s ha e s a ed o oll ou massi e g een
R&D inancing schemes o a (jus ) g een ansi ion. Mos no able a e he passing o
he IRA in he US in 2022 and he announcemen o he EU’s G een Deal Indus ial
Plan in 2023 as pa o he comp ehensi e EU G een Deal. Bo h policy packages a e
subs an ial inancial p og ams o boos he de elopmen and expedi e he deploymen
o clean, emissions- ee echnologies o bo h deca bonize and s eng hen hei espec -
i e economies. While he G een Deal Indus ial Plan is a di ec eac ion o he IRA,
bo h di e conside ably in policy ins umen s, echnology ocus, and expec ed en i on-
men al e ec i eness. Acknowledging he p e iously desc ibed esea ch, we discuss
hese wo g een echnology suppo p og ams wi h espec o hei capaci y o con ib-
u e owa ds an op imal ne -ze o policy mix. We s a wi h a gene al discussion o he
policy mix in he EU e sus he US and hen assess he e ec i eness, e iciency, and
design o he speci ic measu es. This b ings us o some u he speci ic aspec s o op i-
mal policy design, no ye s essed in he p e ious e iew. Be o e doing so, i should
be men ioned ha he IRA has al eady been passed and he unding amoun o
US$370 billion and i s dis ibu ion is speci ied, while o da e, he EU’s G een Deal
Indus ial Plan is mainly an announcemen , whe e only some o he included p og ams
such as he Inno a ion Fund (wo h 40 billion e) o he In es EU P og amme (wo h
26 billion e) as well as REPowe EU (wo h 300 billion e) a e al eady in place. As
such, he EU unding olume is compe i i e wi h he IRA (Fajeau e al. 2023).
The EU and he US di e conside ably in hei gene al clima e policy se ing. The
EU coun ies being pa o Annex B in he Kyo o P o ocol we e among he i s coun-
ies o in e na ionally commi o emission educ ions and join ly o e achie ed hei
a ge s. To achie e his, he EU implemen ed he ETS in 2005, co e ing pa icula ly
he ene gy sec o and ene gy in ensi e indus ies and abou 40% o EU emissions
(ICAP 2023). Al hough accompanied by many o he policy measu es, he EU’s ca bon
p icing scheme is i s main clima e policy measu e.
2
While he EU’s clima e policy
s a egy migh ace c i icism o i s b oad scope, i aligns wi h he ecommended op i-
mal policy by combining a ca bon p ice add essing he en i onmen al ex e nali y and
Jou nal o En i onmen al Planning and Managemen 15
an EI policy o he addi ional ma ke ailu es, such as knowledge ex e nali ies (e.g.
Fische and Newell (2008)).
In compa ison, he US has a ques ionable his o y wi h clima e policy ha ing
signed, hough no a i ied, he Kyo o P o ocol
3
and ha ing a i ied, d opped ou ,
and ejoined he Pa is Ag eemen .
4
Al hough he idea o emission p icing o igina ed
in he US, se e al a emp s o implemen na ional ca bon p icing ailed and ca bon
p icing schemes only exis a he sub-na ional le el (ICAP 2023), co e ing only
6.4% o US GHG emissions in 2021 (OECD 2022). Acknowledging ha ca bon
p icing is less popula among o e s (e.g. Dabla-No is e al. 2023; Dechezlep ^
e e
e al. 2022), he IRA is only a second-bes policy combining en i onmen al, social,
and compe i i eness aspec s in o one policy ool. The esul ing begga - hy-neighbo
policy due o i s ‘Ame ica Fi s ’app oach may lead o ade wa s and c ea e ba -
ie s o o eign ma ke en an s and can educe global wel a e (e.g. Rod ik
(2014)). While he ade implica ions ha e been discussed epea edly (A inasi,
Boeckelmann, and Bap is e 2023), we ocus on he sui abili y o he IRA as an
en i onmen al policy in his pape .
6.1. IRA
The IRA conside s se e al speci ic g een echnologies such as RE, hyd ogen, ca bon
dioxide emo al, and ba e ies. Technology-speci ic suppo ensu es ha a b oad ange
o echnologies a e suppo ed acco ding o hei speci ic needs such ha mo e imma-
u e echnologies a e suppo ed nex o ma u e ones. Al hough he lis o echnologies
is nume ous and b oad, i is no echnology open. Fo example, he IRA gi es ou p o-
duc ion ax c edi s o sola polysilicon o sola wa e bu no o sola hin ilm echnol-
ogy (C edi Suisse 2022). The selec ion o pa icula echnologies ma ches he conce n
abou go e nmen s’limi ed abili y in picking winne s (Acemoglu e al. 2016). This is
mainly because i excludes no el app oaches and is p one o esul in an ine icien
echnology mix since, o example, oday’s mos e icien echnologies may no be he
mos e icien in he u u e depending on u he oppo uni ies o lea ning and echno-
logical imp o emen s (Lancke and Quaas 2019). The e o e, he IRA can induce
echnological lock-in and pa h dependencies (Ce an es e al. 2023). This is e en mo e
p oblema ic as ca bon p icing as a gene al ma ke -push ins umen is missing (Fische
and Newell 2008).
The ype o echnology add essed is also in luenced by he choice o suppo
ins umen s. In Ame ican policy adi ion, he IRA is mainly p o iding ax c edi s o
p oduc ion and in es men . A imes, hese a e combined wi h a compe i i e bid o
ensu e suppo o he mos e icien echnology. G an s only play a mino ole.
Al hough bo h ax c edi s and g an s a ge he same goal –inc easing he in es men
in g een R&D – hey dis inguish hemsel es by he iming o he paymen . G an s
imply an up on paymen while ax c edi s a e only ecei ed a e p oduc ion and
in es men . Wi h ax c edi s he isk o go e nmen s is low and he payback ime is
sho (Ce an es e al. 2023), while i ms s ill ace he b un o he isk associa ed
wi h R&D in es men s. Thus, he e is a isk ha he IRA canno ully add ess he ma -
ke ailu e ela ed o knowledge c ea ion (see Sec ion 2.1). In addi ion, and as dis-
cussed in Sec ion 4, he e is empi ical e idence o he go e nmen al suppo o RE
ha he e ec i eness o induce inno a ion is lowe unde ax c edi s in compa ison o
g an s (Hille, Al hamme , and Diede ich 2020). Going beyond his, i ms need o incu
16 L. P. Meissne e al.
axes o ecei e ax c edi s, which equi es a ma ke able and p o i able p oduc –
al hough his is pa ially ci cum en ed by allowing o ans e s (The Whi e House
2023). The e o e, ax c edi s by design p eselec ma u e, deployable p oduc s and
exclude nascen echnologies. The p eselec ion may be bene icial o push exis ing ech-
nologies and eap he low-hanging ui s in emission educ ions, bu he lack o sup-
po o inc emen al inno a ion can ha e long- e m e ec s on he en i onmen , as
u he de elopmen s o imma u e and no el EIs will be necessa y o ne -ze o emis-
sions (In e na ional Ene gy Agency 2021). Fu he mo e, Roy, Bu aw, and Renne
(2021) demons a e ha ax c edi s achie e lowe emission educ ions a a highe cos
in compa ison o ca bon p icing, u he ques ioning he sui abili y o ax c edi s as a
second-bes en i onmen al policy.
The e is no only he ques ion o which echnologies a e suppo ed bu also how
much suppo a ce ain echnology ecei es. E iciency would wa an ha all emission
educ ions om equally ma u e echnologies ecei e he same subsidy paymen and
compe i i eness will selec he winne . Howe e , his is no he case in he IRA. Fo
example, g een hyd ogen om elec olysis can ecei e a highe p oduc ion ax c edi
han blue hyd ogen ( om ossil uels wi h ca bon cap u e and s o age) as a hyd ogen
p oduc ion ax c edi can be combined wi h an RE p oduc ion ax c edi bu no wi h
ca bon cap u e and s o age p oduc ion ax c edi (C edi Suisse 2022). Ye , he ne
emissions a e he same and bo h g een and blue hyd ogen ha e u he en i onmen al
d awbacks ha make weighing hem di icul (hea y eliance on eshwa e and isk o
ca bon leakage, espec i ely). Acco dingly, such di e en ia ion is no wa an ed om
an e iciency pe spec i e and, ins ead, h ea ens echnological lock-in. This could be
de imen al o he long e m en i onmen al objec i e.
Fu he mo e, he IRA is cons uc ed in a way ha he e is a base c edi amoun
ha can be inc eased wi h bonus c edi amoun s when ce ain c i e ia a e me . Fo
example, he p oduc ion ax c edi o RE has a baseline a e o 0.3 cen s/kW and can
be inc eased i e old when he p ojec mee s a p e ailing wage equi emen , can be
inc eased by 10% i domes ic p oduc s a e used in he manu ac u ing p ocess o i
loca ed in an ene gy communi y (The Whi e House 2023). Since he IRA a ge s en i -
onmen al, compe i i eness and social domains wi h one policy ins umen – hus, kill-
ing many bi ds wi h one s one, i is p one o cos ine iciencies (Fische , H€
uble , and
Schenke 2021; Tinbe gen 1952). Fi s , using a single policy ins umen may blu he
lines be ween he di e en ma ke ailu es. Second, p o iding bonus c edi s is e icien
i c edi amoun s ha e been se op imally. High c edi amoun s a e necessa y o coun-
e he inexis ence o a ca bon p ice. Ne e heless, his is solely con ingen on he
p esence o a ma ke ailu e; i he e is no unde lying ma ke ailu e, he e is also no
eason o a bonus c edi . Compe i i eness is no associa ed wi h any ma ke ailu e
and, hence, an inc ease in he ax c edi amoun o RE p oduc ion based on local con-
en equi emen s is no wa an ed. Such mixing o indus ial policy a ge s wi h en i -
onmen al policy unnecessa ily inc eases he cos and ine iciency o he en i onmen al
policy o inc ease RE.
Mo eo e , wi h he abili y o i ms o s ack a ious c edi s oge he , he e iciency
o he IRA is pu in ques ion. Fo example, i a p oduce o emission- ee a ia ion uel
can secu e a p oduc ion c edi o p oducing such uel, c edi s o p oducing RE o i s
manu ac u ing p ocess, o using ca bon cap u e and s o age in i s p ocess, and o
using domes ic p oduc s, he p oduc ion p ocess may no be chosen o e iciency ea-
sons bu o inancial easons by choosing he p ocess by which he mos ax c edi s
Jou nal o En i onmen al Planning and Managemen 17
can be eaped. Addi ionally, his can lead o echnological lock-in, h ea ening he
IRA’s abili y o achie e emission educ ions in he u u e.
In summa y, wi h i s ocus on ax c edi s, he IRA can inc ease he compe i i eness
o ma u e echnologies and accele a e hei deploymen . Fo he same easons, how-
e e , he IRA’s success in achie ing emission educ ions is likely cons ained o he
sho e m by ocusing on a se o speci ic and mo e ma u e echnologies. Mo eo e ,
ax c edi a es may become ine icien due o he use o a single policy ins umen o
add ess di e en ma ke ailu es, h ough he possibili y o inc easing ax a es o
non- echnological s anda ds, and due o he abili y o s ack a ious c edi a es. I is
also ques ionable whe he he IRA can achie e ac ual emission educ ions bo h in he
sho and long e m, no only due o i s ocus on exis ing echnologies bu also due o
possible ebound e ec s (Sec ion 5) ha become likely because o missing comple-
men a y ca bon p icing. Al hough he na ional success o he IRA is ques ionable, he
hea y subsidiza ion o he IRA may lead o global wel a e imp o emen s by educing
he cos o EIs, inc easing hei ake-up globally, and hus po en ially educing global
emissions.
6.2. EU G een Deal
Compa ed o he IRA, he EU G een Deal Indus ial Plan o esees a g ea e a ie y o
ins umen s and echnologies. The EU ocuses on g an s as well as public p ocu emen
s a egies, compe i i e auc ions, bu also ax c edi s. As ‘one size does no i all’(see
Pi elis, Vasilakos, and Chal a zis 2020), a ious policy ins umen s o di e en ech-
nologies and sec o s can a ge echnologies o a ious ma u i y le els. Al hough he
EU G een Deal Indus ial Plan ocuses on he same echnologies as he IRA i is less
echnologically speci ic. Fo example, he Inno a ion Fund p omo es any p ojec ha
can lead o signi ican emission educ ions (Eu opean Commission 2023b) and he eby,
does no single ou any speci ic echnologies. Howe e , he Inno a ion Fund suppo s
only p ojec s ha a e ‘su icien ly ma u e in e ms o planning, business model and
inancial and legal s uc u e’.
5
The e o e, he G een Deal Indus ial Plan can cap u e a
b oad spec um o echnologies and is open o no el clean echnologies.
Al hough he de ails o he EU G een Deal Indus ial Plan a e no ye clea , i
pledges o adjus subsidy paymen s o hose o o he coun ies (e.g. he US) o emain
compe i i e and a oid EU i ms eloca ing (Eu opean Commission 2023a). Thus, i he
IRA speci ies a ce ain subsidy le el ha is g ea e han wha can be ecei ed ia an
EU unding p og am, he EU ensu es ha domes ic i ms ecei e he same amoun .
While his may be impo an o le el he playing ield wi h he US, such a pledge may
no only uel a subsidy wa (Inagaki, Chazan, and Fleming 2023) bu i also means
ha possible ine iciencies o he IRA will be eplica ed in he EU. Fu he mo e, i
igno es ha , by design, he IRA needs o employ highe suppo olumes o achie e
gi en emission a ge s since na ional emission p icing is missing. Ul ima ely he o e -
all cos o clima e policy in he EU would subs an ially inc ease and de ia e om he
i s -bes policy mix.
The EU’s mul i ace ed app oach has a g ea e po en ial o educe emissions while
also being mo e e icien . The EU G een Deal Indus ial Plan co e s an a ay o policy
ins umen s and echnologies in a ious s ages o de elopmen and is embedded in an
a ay o o he policy and unding schemes and mos impo an ly accompanied by a
s ong ca bon p icing scheme. By se ing, i can be bo h cos -e icien and e ec i e in
18 L. P. Meissne e al.
educing emissions. Ne e heless, especially he IRA bu also he EU’s G een Deal
Indus ial Plan p io i ize he deploymen o clean echnologies o indus ial policy ea-
sons as an a emp o ha e a po en ial i s -mo e ad an age and gain aluable ma ke
sha es in he clean ene gy ma ke wo h billions (In e na ional Ene gy Agency 2023).
Howe e , i emains ques ionable whe he a ocus on ma u e clean echnologies will
be he winning s a egy o coun ies o cap u e he lion’s sha e o he g een indus y
o whe he coun ies will maneu e hemsel es in o echnological lock-in.
Ne e heless, he global communi y may bene i om he push o EI close o he
ma ke .
7. Conclusion and esea ch ou looks
Ou e iew shows ha go e nmen al esea ch and de elopmen (R&D) suppo o
en i onmen al inno a ion (EI) should be pa o an e ec i e and e icien policy mix
o achie e ne -ze o emission a ge s. I is well-es ablished, in heo y, and eplica ed in
empi ics, ha he op imal policy mix combines an en i onmen al policy, in pa icula
emission p icing, wi h an inno a ion policy o co e he win-ma ke ailu es. O e all,
a policy mix os e s he de elopmen and deploymen o EI, signi ican ly educes he
cos o achie ing gi en emissions a ge s, and ul ima ely educes emissions e ec i ely.
As a s andalone policy, nei he emission p icing no an R&D policy can e icien ly
a ge ma ke ailu es. In heo y, se e al ma ke ailu es, such as non-in e nalized
knowledge c ea ion, dynamic e u ns o EIs and pa h dependency, jus i y go e nmen al
R&D suppo . In addi ion, empi ical esul s show ha R&D suppo inc eases EI ac i -
i ies, suppo s he deploymen o clean echnologies and educes emissions. Howe e ,
R&D suppo is signi ican ly less e icien as a s andalone policy compa ed o emis-
sion p icing. None heless, inno a ion policies a e widely adop ed as second-bes en i -
onmen al policy –as is he case wi h he US IRA. The obse a ion ha R&D policy is
mo e publicly accep ed and poli ically easible han a i s -bes policy mix explains
his bias owa ds inno a ion policy.
When designing public R&D suppo o EI, in e en ion is especially
ele an o imma u e clean echnologies, whe e dynamic e u ns, o ins ance h ough
lea ning-by-doing, can signi ican ly educe p oduc ion cos s. This calls o di e en i-
a ed echnology suppo , e en hough ocusing on s a ic cos s only, would imply ha
each echnology ecei es he same suppo pe uni o emission educ ion. A he same
ime, echnology open suppo is necessa y o a oid go e nmen ’s ailu e in picking
winne s, which is no o iously di icul due o he dynamic e u ns and pa h
dependency.
Based on he insigh s o he li e a u e e iew, we discussed and compa ed bo h he
US’s IRA and he EU G een Deal Indus ial Plan in hei abili y o suppo EI and
achie e emission educ ions. T ying o achie e en i onmen al, social, and compe i i e
goals in a single policy, he IRA is ine ec i e in achie ing he a ious a ge s. The
use o ax c edi s and he ocus on exis ing echnologies can achie e emission educ-
ions in he sho e m by pushing he deploymen o ma u e echnologies and helping
he US o secu e a i s -mo e ad an age. None heless, i lacks he long- e m pe spec-
i e by picking winne s ea ly on and by dis ega ding he impo ance o imma u e inno-
a ions o ne -ze o emissions.
In compa ison, he EU G een Deal Indus ial Plan can e ec i ely complemen he
exis ing ca bon p icing scheme o he EU and achie e a heo e ically i s -bes policy
Jou nal o En i onmen al Planning and Managemen 19
mix, educing he cos o achie ing clima e a ge s. The EU G een Deal Indus ial
Plan in ends o combine a ious exis ing schemes in o a comp ehensi e scheme o
push EI. The eby, i allows o g ea e echnological openness han he IRA h ough a
ocus on he same co e, ma u e echnologies. Howe e , i s plan o inc ease subsidy
le els o equal hose o he IRA is ine icien , as he equi ed suppo le els o he EU
a e lowe because he EU mus no co e he en i onmen al pe spec i e as he IRA
has o. Such ac ions only inc ease he cos o en i onmen al policies o he EU and
add uel o a po en ial subsidy wa .
While we a gue ha hese insigh s include aluable lessons o decision ake s, we
acknowledge ha mo e speci ic policy ecommenda ions a e di icul o de i e. The
indings a e ypically ei he heo e ical o linked o e y speci ic se ings and hus do
no allow, o example, o say much abou which echnology o suppo h ough which
exac measu e and wi h which amoun . So when, e.g. looking a he EU G een Deal
wi h i s se e al echnology suppo p og ams o he US In la ion Reduc ion Ac , one
can say only e y gene ally whe he his is in line wi h an op imal policy mix (in gen-
e al yes o he EU and no o he US), po en ially leading o ine iciencies (p obably
bo h, bu mo e so o he IRA) o which he main con ibu ion owa ds deca boniza ion
is (lowe cos s o ache ing gi en a ge s in he EU, achie ing emission educ ions a
all in he US). In addi ion, many di e en o ms o R&D policy exis , so i is no
homogeneous i sel . The discussion o he design o an op imal R&D policy in he
en i onmen al policy mix goes beyond he scope o his pape bu opens he doo o
u u e esea ch.
In addi ion, we only ouched upon ano he in e es ing global pa e n, which gi es
oom o u u e esea ch. The e is an ongoing deba e abou os e ing na ional indus ies
and hei compe i i eness besides e icien ly and e ec i ely achie ing emission a ge s
as addi ional a ge s o R&D suppo measu es. We ha e only ouched upon his issue,
since i is no he ocus o his pape and ela es o o he s ands o li e a u e dealing,
e.g. wi h s a egic ade policy o indus ial policy. Acknowledging ha domes ic wel-
a e is a a ge o poli icians o which hey a e mos likely elec ed, assesses an op i-
mal R&D suppo p og am o be e en mo e di icul . Finally, al hough we conside a
as numbe o s udies o de i e insigh s on go e nmen al R&D suppo and EIs, we
ha e no implemen ed a me a-analysis. When zooming in on speci ic domains o R&D
suppo o EIs, a me a-analysis could p o ide u he insigh s.
O e all, i will emain impo an o inc easingly e alua e speci ic public R&D p o-
g ams o lea n mo e abou wha makes hem success ul in en i onmen al and eco-
nomic e ms. In line wi h his, we emphasize se e al po en ial u u e esea ch
di ec ions.
Fi s , R&D expendi u e is a b oadly de ined a iable in many empi ical
s udies. Depending on he s udy, he a iable cap u es any hing om gene al na ional
(go e nmen and i m) R&D budge s o go e nmen al R&D suppo o speci ic g een
echnologies. In many cases, i is di icul o he eade o unde s and wha go e n-
men al R&D measu es, e.g. whe he an induced esea ch en i onmen o a ge ed sup-
po o speci ic echnologies. To cap u e he ue essence o go e nmen al R&D
suppo o EI, mo e ine g ain da a on he ype o go e nmen al R&D suppo would
be powe ul. This is especially ele an conside ing he mul iple ex e nali ies, such as
spillo e s, ha can s ee he esul s away om i s a ge . Building on his ou lined
esea ch gap, popula da ase s used in empi ical s udies ela ed o i ms’inno a ion
ac i i ies, such as he Communi y Inno a ion Su ey, do no dis inguish be ween
20 L. P. Meissne e al.
gene al subsidies and subsidies a ge ed owa ds imp o ing i ms’en i onmen al pe -
o mance. Conside ing he a ge o a subsidy allows esea che s o analyze how go -
e nmen al R&D a ge ed o en i onmen al ou comes ela es o i ms’en i onmen al
pe o mance.
Second, mos empi ical s udies conside p i a e R&D expendi u e a he han go -
e nmen al R&D suppo when analyzing he impac on emission educ ion. Fu u e
s udies migh ocus on how go e nmen al R&D suppo can le e age p i a e R&D
which helps i ms o imp o e hei en i onmen al pe o mance. Such esea ch can also
help o be e unde s and he channel on how go e nmen al R&D suppo can lead o
educed en i onmen al deg ada ion. In his ligh , i is also o high in e es o analyze
whe he go e nmen al R&D suppo induces addi ional R&D ac i i ies o shi s he
inancial cos s om p i a e ac o s o go e nmen s.
Thi d, empi ical assessmen s o go e nmen al R&D suppo a e p ima ily limi ed o
RE echnologies and hus, he e is a selec ion bias ocused on echnology winne s. Le ing
he echnology winne s w i e his o y may o e es ima e he e ec i eness o public R&D on
inno a ion. Fu u e empi ical esea ch should b oaden he scope o EIs co e ed.
Fou h, i is impo an o conside he egula o y en i onmen when assessing how
go e nmen al R&D ela es o i ms’emission educ ions. In pa icula , he
ele ance o he policy mix is discussed in de ail in he li e a u e e iew. Acco dingly,
esea che s should conside whe he a i m is also exposed o ma ke - ela ed o o he
non-ma ke - ela ed en i onmen al policies and can analyze how his ela es o he
impac o g een R&D suppo . In so doing, using c oss-coun y da a could be powe -
ul. In he bes case, he da a should co e coun ies in di e en egions a di e en
s ages o de elopmen .
Fi h, howe e , empi ical analysis using single-coun y da a can also signi ican ly
con ibu e o ou unde s anding o how go e nmen al R&D can induce emission educ-
ions. Applying s a e-o - he-a econome ic echniques, such as ecen de elopmen s in
he di e ence-in-di e ence li e a u e, can help o e eal a causal impac o a speci ic
R&D policy. Al hough i migh ace limi ed ex e nal alidi y, such a nuanced e en s udy
can es ima e he impac o a speci ic policy on i ms’en i onmen al pe o mance and
migh also help o unde s and how di e en o ms o go e nmen al R&D suppo EIs.
Finally, i emains impo an o u he dis inguish be ween he di e en le els o
ma u i y o EIs. As poin ed ou –R&D policy is pa icula ly ele an o EIs in he ea ly
s ages o ma u i y. Howe e , o such echnologies, he impac on emission educ ions
can occu wi h a subs an ial ime lag and he eal po en ial can un old wi h he di usion
a a la e s age o ma u i y. In his case, he ue impac will no be obse able o he di -
ec ly suppo ed i m. Ins ead, o he i ms –e en beyond he bo de –migh imp o e hei
en i onmen al pe o mance. So a , he e is no much known abou how na ional R&D
suppo leads o emission educ ions in o he coun ies a a la e poin in ime.
In summa y, we hope ha he comp ehensi e li e a u e e iew and he possible
u u e esea ch di ec ions ou lined can p o ide aluable insigh s o s ee o hcoming
heo e ical and empi ical s udies conce ning he ole o go e nmen al R&D suppo in
accele a ing he ake-o o EIs.
No es
1. Fo u he in o ma ion on he Eu opean G een Deal isi : h ps://commission.eu opa.eu/
s a egy-and-policy/p io i ies-2019-2024/eu opean-g een-deal_en, accessed on 25 July 2023.
Jou nal o En i onmen al Planning and Managemen 21
2. Fo mo e in o ma ion on he EU emission ading scheme isi h ps://clima e.ec.eu opa.eu/
eu-ac ion/eu-emissions- ading-sys em-eu-e s_en, accessed 4 Augus 2023.
3. See he Uni ed S a es a chi es o mo e in o ma ion h ps://1997-2001.s a e.go /global/
global_issues/clima e/ s-us_sign_kyo o_981112.h ml, accessed 4 Augus 2023.
4. See s a emen by he Uni es S a es go e nmen h ps://www.s a e.go / he-uni ed-s a es-
o icially- ejoins- he-pa is-ag eemen /, accessed 4 Augus 2023.
5. Fo mo e in o ma ion isi h ps://clima e.ec.eu opa.eu/eu-ac ion/ unding-clima e-ac ion/
inno a ion- und/wha -inno a ion- und_en, accessed 3 Augus 2023.
Acknowledgemen s
A p e ious e sion has been p esen ed a he Kiel Ins i u e Wo kshop on ‘The ole o public
esea ch and inno a ion measu es on mi iga ing clima e change’2023. In addi ion, we would
like o hank Ma c Blaue , Di k Dohse, Thomas Heime , Hen ie e Zoe Figue oa Agos o,
Raphael Be endsohn and wo anonymous e e ees o hei suppo and aluable eedback.
Disclosu e s a emen
The au ho s decla e ha hey ha e no known compe ing inancial in e es s o pe sonal
ela ionships ha could ha e appea ed o in luence he wo k epo ed in his pape .
Funding
We acknowledge inancial suppo by he Fede al Minis y o Educa ion and Resea ch unde he
g an numbe [01LA2101A].
ORCID
Leonie P. Meissne h p://o cid.o g/0009-0002-8889-2869
Sonja Pe e son h p://o cid.o g/0000-0002-7379-2681
Finn Ole Sem au h p://o cid.o g/0000-0002-6894-2577
Re e ences
Acemoglu, Da on, Philippe Aghion, Leona do Bu sz yn, and Da id Hemous. 2012. “The
En i onmen and Di ec ed Technical Change.”The Ame ican Economic Re iew 102 (1):
131–166. doi:10.1257/ae .102.1.131.
Acemoglu, Da on, U uk Akcigi , Douglas Hanley, and William Ke . 2016. “T ansi ion o Clean
Technology.”Jou nal o Poli ical Economy 124 (1): 52–104. doi:10.1086/684511.
Aghion, Philippe, An oine Dechezlep ^
e e, Da id H
emous, Ral Ma in, and Reenen Van John.
2016. “Ca bon Taxes, Pa h Dependency, and Di ec ed Technical Change: E idence om he
Au o Indus y.”Jou nal o Poli ical Economy 124 (1): 1–51. doi:10.1086/684581.
Aghion, Philippe, Came on Hepbu n, Alexande Tey elboym, and Dimi i Zenghelis. 2019. “Pa h
Dependence, Inno a ion and he Economics o Clima e Change.”In Handbook on G een
G ow h,67–83. Chel enham: Edwa d Elga Publishing. doi:10.4337/9781788110686.00011.
Alam, Md Samsul, Nicholas Ape gis, Sudha shan Reddy Pa ama i, and Jianchun Fang. 2021.
“The Impac s o R&D In es men and S ock Ma ke s on Clean-Ene gy Consump ion and
CO
2
Emissions in OECD Economies.”In e na ional Jou nal o Finance & Economics 26
(4): 4979–4992. doi:10.1002/ij e.2049.
Alam, Md Samsul, Muhammad A i , Chu Chien-Chi, and U
gu Soy as¸. 2019. “Does Co po a e
R&D In es men A ec Fi m En i onmen al Pe o mance? E idence om G-6 Coun ies.”
Ene gy Economics 78: 401–411. doi:10.1016/j.eneco.2018.11.031.
22 L. P. Meissne e al.
Ang, Ge aldine, Di k R€
o ge s, and P alhad Bu li. 2017. The Empi ics o Enabling In es men
and Inno a ion in Renewable Ene gy. Pa is: OECD. doi:10.1787/67d221b8-en.
A inasi, Ma ia G azia, Lukas Boeckelmann, and Meunie Bap is e. 2023. “Un iendly F iends:
T ade and Reloca ion E ec s o he US In la ion Reduc ion Ac .”CEPR.h ps://cep .o g/
oxeu/columns/un iendly- iends- ade-and- eloca ion-e ec s-us-in la ion- educ ion-ac .
Be nd , E ns R., Robe S. Pindyck, and Pie e Azoulay. 2003. “Consump ion Ex e nali ies and
Di usion in Pha maceu ical Ma ke s: An iulce D ugs.”The Jou nal o Indus ial
Economics 51 (2): 243–270. doi:10.1111/1467-6451.00200.
Be am, Ch is oph, Gunna Lude e , Robe C. Pie zcke , E a Schmid, Elma K iegle , and
O ma Edenho e . 2015. “Complemen ing Ca bon P ices wi h Technology Policies o Keep
Clima e Ta ge s wi hin Reach.”Na u e Clima e Change 5 (3): 235–239. doi:10.1038/
nclima e2514.
B€
oh inge , Ch is oph, And eas Kelle , Ma kus Bo olamedi, and Anelise Rahmeie Sey a h. 2016.
“Good Things Do No Always Come in Th ees: On he Excess Cos o O e lapping Regula ion
in EU Clima e Policy.”Ene gy Policy 94: 502–508. doi:10.1016/j.enpol.2015.12.034.
Bond, S ephen, Die ma Ha ho , and John Van Reenen. 2005. “In es men , R&D and Financial
Cons ain s in B i ain and Ge many.”Annales D’
Economie E De S a is ique. Genes,
ADRES 80 (79): 433–460. doi:10.2307/20777584.
B ockway, Paul E., S e e So ell, G ego Semieniuk, Ma hew Kupe us Heun, and Vic o Cou .
2021. “Ene gy E iciency and Economy-Wide Rebound E ec s: A Re iew o he E idence
and I s Implica ions.”Renewable and Sus ainable Ene gy Re iews 141 (May): 110781. doi:
10.1016/j. se .2021.110781.
B ucal, A lan, Bea a Ja o cik, and Inessa Lo e. 2019. “Good o he En i onmen , Good o
Business: Fo eign Acquisi ions and Ene gy In ensi y.”Jou nal o In e na ional Economics
121: 103247. doi:10.1016/j.jin eco.2019.07.002.
Ca ley, Sanya. 2009. “S a e Renewable Ene gy Elec ici y Policies: An Empi ical E alua ion o
E ec i eness.”Ene gy Policy 37 (8): 3071–3081. doi:10.1016/j.enpol.2009.03.062.
Ce an es, Ma io, Chia a C iscuolo, An oine Dechezlep ^
e e, and Di k Pila . 2023. D i ing Low-
Ca bon Inno a ions o Clima e Neu ali y. Pa is: OECD. doi:10.1787/8e6ae16b-en.
Chen, Fenglong, Meichang Wang, and Zhengning Pu. 2022. “The Impac o Technological
Inno a ion on Ai Pollu ion: Fi m-Le el E idence om China.”Technological Fo ecas ing
and Social Change 177: 121521. doi:10.1016/j. ech o e.2022.121521.
Copeland, B ian R., Joseph S. Shapi o, and M. Sco Taylo . 2022. “Globaliza ion and he
En i onmen .”In Vol. 5 o Handbook o In e na ional Economics, edi ed by Gi a Gopina h,
Elhanan Helpman, and Kenne h Rogo , 61–146. Ams e dam: Else ie . doi:10.1016/bs.
hesin .2022.02.002.
Cos an ini, Vale ia, F ancesco C espi, Chia a Ma ini, and Luca Pennacchio. 2015. “Demand-
Pull and Technology-Push Public Suppo o Eco-Inno a ion: The Case o he Bio uels
Sec o .”Resea ch Policy 44 (3): 577–595. doi:10.1016/j. espol.2014.12.011.
Cos an ini, Vale ia, F ancesco C espi, and Alessand o Palma. 2017. “Cha ac e izing he Policy
Mix and I s Impac on Eco-Inno a ion: A Pa en Analysis o Ene gy-E icien
Technologies.”Resea ch Policy 46 (4): 799–819. doi:10.1016/j. espol.2017.02.004.
C edi Suisse. 2022. “US In la ion Reduc ion Ac : A Ca alys o Clima e Ac ion.”h ps://www.
c edi -suisse.com/abou -us-news/en/a icles/news-and-expe ise/us-in la ion- educ ion-ac -a-
ca alys - o -clima e-ac ion-202211.h ml.
Dabla-No is, E a., Thomas Helbling, Salma Khalid, Hibah Khan, Giacomo Magis e i,
Alexand e Sollaci, and K ishna S ini asan. 2023. Public Pe cep ions o Clima e Mi iga ion
Policies: E idence om C oss-Coun y Su eys. Washing on: In e na ional Mone a y Fund.
h ps://www.im .o g/en/Publica ions/S a -Discussion-No es/Issues/2023/02/07/Public-
Pe cep ions-o -Clima e-Mi iga ion-Policies-E idence- om-C oss-Coun y-Su eys-528057.
De Loecke , Jan. 2007. “Do Expo s Gene a e Highe P oduc i i y? E idence om Slo enia.”
Jou nal o In e na ional Economics 73 (1): 69–98. doi:10.1016/j.jin eco.2007.03.003.
Dechezlep ^
e e, An oine, Ad ien Fab e, Tobias K use, Bluebe y Plan e ose, Ana Sanchez Chico,
and S e anie S an che a. 2022. Figh ing Clima e Change: In e na ional A i udes owa d
Clima e Policies. Pa is: OECD. doi:10.1787/3406 29a-en.
Eu opean Commission. 2023a. “Communica ion: A G een Deal Indus ial Plan o he Ne -Ze o
Age.”h ps://commission.eu opa.eu/documen /41514677-9598-4d89-a572-abe21cb037 4_en.
Jou nal o En i onmen al Planning and Managemen 23
